Repeat Order Probability Report

When is the Incremental Event Probability perspective available?

The incremental event probability perspective is only available when filters use dimensions that are equal for all orders (for example, user’s gender, user’s age or user’s source).

This is because this perspective relies on a dimension called User's order number for segmentation, which numbers a user’s purchases (for example, John’s 1st, 2nd, and 3rd orders).

If you added a filter that uses a dimension which is not equal for all orders (for example, Order's Region), the User's order number dimension would no longer be accurate. This is because it does not account for specific regions when numbering a user’s orders (for example, John’s 1st, 2nd, 3rd orders are still the same, regardless of their region).

Turning an order-specific dimension into a user-specific dimension

In certain cases, you may be able to turn an order-specific dimension into a user-specific dimension to add as filter in the Repeat Order Probability chart. In these cases, you return the order attribute of a user’s first order or latest order (for example, User’s first order region name).

If you want to create such a new dimension, contact support.

Comparing repeat probability of orders with different attributes

To compare the number of repeat purchases for different order attributes (for example, order’s region), Adobe recommends creating a chart similar to Users by lifetime number of orders. This shows you the number of users that made 1, 2, 3,… lifetime number of orders, and add the order level filter. (in other words, This can show you whether users make more or less repeat purchases in one region or another.)

The numbers that make up such a chart can then be exported to excel to calculate the repeat order probability ratio. To see the probability of customers that made (x) orders to make (x+1) orders, simply divide the number of people who've made at least (x+1) purchases by the number of people who have made at least (x) purchases.

Example: